This project delves into the analysis of a loan default dataset. Its primary objectives include identifying the key factors influencing loan defaults, exploring customer segments with the highest default risk, and constructing a predictive model to identify high-risk clients.

• Source Data Folder: Contains raw source datasets used in the analysis.
• Derived Data Folder: Contains processed datasets generated during data transformation and cleaning.
• Notebooks:
  • Data_Preprocessing.ipynb: Notebook containing data preprocessing steps and transformations.
  • EDA.ipynb: Exploratory Data Analysis (EDA) notebook with data visualization and initial insights.
  • Data_processing.ipynb: Notebook for data processing to prepare data for modeling
  • Modeling.ipynb: Notebook for building and evaluating machine learning models.
• Utils Package: Contains utility functions used in modeling.
• Results: Contains presentation of results.

1. Clone the repository:

   https://github.com/ManukyanRazmik/Civitta_project.git
   

2. Install requirements.txt

    pip -r install requirements.txt
   

3. Run 1_Data_preprocessing.ipynb
4. Run 2_EDA.ipynb
5. Run 3_Data_processing.ipynb
6. Run 4_Modeling.ipynb